Network File System (NFS) servers are typically used to host, among other things, boot images for a large number of virtual machines. Remote Procedure Calls (RPC) are periodically made by a monitoring agent to determine health or state of the NFS server. However, during a boot-storm, i.e., when hundreds of virtual machines are booting up at the same time from an NFS server, RPC calls may not adequately monitor the state of the NFS server because the NFS server may be overloaded. For example, the RPC calls may time out and report the NFS server as faulted. In addition, these RPC calls may add to the existing load at the NFS server. As a result, monitoring a server using traditional RPC calls may not be entirely accurate, efficient, or reliable.
In view of the foregoing, it may be understood that there may be significant problems and shortcomings associated with current monitoring of server health.